Bill Thurston cited personal communication with himself, which is a pretty unusual source.

I will reject a paper if it references a private communication with author oneself. Just tell the truth, undisclosed prior work.

I believe it was his thesis and I assume it was tongue in cheek.

Inevitably, I’m going to have to ask for a reference.

He used to sit on the floor of his office with the lights off under a giant moving box and brainstorm math.

Reference: personal communication

Amusingly enough, it's personal communication with one of his students.

Ethically do you even have to cite something properly when the person who solved it was intentionally anonymous? If the concern is that people will think the person writing solved it, they can just write a sentence where it came from.

Academia has no problem with pseudonymous publication. See "Student" of "Student's T distribution" fame. [1]

[1] https://en.m.wikipedia.org/wiki/William_Sealy_Gosset

> Academia has no problem with pseudonymous publication

another great example is Bitcoin paper.

> Ethically do you even have to cite something properly when the person who solved it was intentionally anonymous?

Yes. Not because they need/want the credit, but because it's important to note that it's not your idea.

> If the concern is that people will think the person writing solved it, they can just write a sentence where it came from.

That is, basically, what a citation is.

Not to mention that a citation is there not only to show the world whose content you are referencing, but also because in case the reader doesn't believe you and wants to go check, they can do so by following the citation

Who can proof that it isn't your idea?

Intentional academic dishonesty is an entirely separate issue. If you're going to plagiarize, it doesn't really matter if there's a formal way of citing the material you don't plan to cite.

4chan

citing "anonymous", if it is unclear who the author is, is nothing new. I did that a long time ago when I a cited something from the nmap man page in a paper for university . Back then it was unclear - at least to me - who the author is. This is a very common practice. I do not see, where the novelty is.

Nowadays you could simply check with git blame right?

if it was just a pseudonym, one may still cite it as anonymous.

I guess we cite the poster's hash and timestamp?

What do you mean by hash? 4chan posts use an incrementing id per board, so if you put >>>/sci/3751197 you have a unique identifier for the post (sci being the board and 3751197 being the post number). For convenience you could link to an archive site as well, but they're known to die every few years.

Some boards like /pol/ have a user hash, unique per user-session, stable across posts.

The hashes assigned to your IP on the boards that have them are per-thread to my knowledge. All your posts in one thread will have a particular hash, but your post in another thread will have a different one.

Ah I forgot about that. It would certainly be a good idea to include it if available. Do you know if it is necessary for a unique id? I've never browsed any of the boards with user hashes.

Not sure what you mean by necessary. Post id would still be unique.

You'd need a link for people to see the original content too so you'd want to include a link to one of the archive sites I guess (as a second best source for an ephemeral post). It's odd but there are citation patterns for websites, wonder if they include a standard for including archives of sources not available at time of writing.